Commits · 244e5b2f79399fe7f1f0bab9077532b01e0edfb3 · general / dsml / emoUS-public

Jun 21, 2022
- add wow downstream · 244e5b2f
  zqwerty authored 2 years ago
  
  244e5b2f
Jun 20, 2022
- add woz dataset · a323f373
  zqwerty authored 2 years ago
  
  a323f373
- add woz dataset · c3167c9d
  zqwerty authored 2 years ago
  
  c3167c9d
Jun 02, 2022
- add wow data · aede1333
  zqwerty authored 2 years ago
  
  aede1333
May 30, 2022
- update pre-training data · ec2167a3
  zqwerty authored 3 years ago
  
  ec2167a3
May 27, 2022
- update dailydialog: add space after full-stop · 4a5a0d4b
  zqwerty authored 3 years ago
  
  4a5a0d4b
- key2gen_noisy: 1) rg; 2) keywords grounded generation; 3) add noisy keywords... · 5ead12a1
  zqwerty authored 3 years ago
  5ead12a1
May 17, 2022
- add persona chat datasets · 243040d1
  zqwerty authored 3 years ago
  
  243040d1
- check.py allow absent field: domain, goal, da, state, db · d24454a7
  zqwerty authored 3 years ago
  
  d24454a7
May 16, 2022
- update kvret: change the order of (slot, value) in db entry · bf9dd000
  zqwerty authored 3 years ago
  
  bf9dd000
May 13, 2022
- add kvret · 82f0b287
  zqwerty authored 3 years ago
  
  82f0b287
May 12, 2022
- update dart testset · 7f5072d6
  zqwerty authored 3 years ago
  
  7f5072d6
May 11, 2022
- add commongen dataset · 7b73fe6d
  zqwerty authored 3 years ago
  
  7b73fe6d
- add key2gen tasks:nlg&dart · cfe30c75
  zqwerty authored 3 years ago
  
  cfe30c75
- run_seq2seq will use early stop(patient=10 epochs) when using load_best_model_at_end param · 96c469b9
  zqwerty authored 3 years ago
  
  96c469b9
- add dart dataset · a403317b
  zqwerty authored 3 years ago
  
  a403317b
- add key2gen metric. pre-training dataset add dailydialog · e265e86b
  zqwerty authored 3 years ago
  
  e265e86b
May 07, 2022
- update keywords extraction: remove punctunation and do not merge tokens into a... · a1a2f240
  zqwerty authored 3 years ago
  a1a2f240
- add rm datadir before multiple datasets training to avoid repeating samples · beda1b76
  zqwerty authored 3 years ago
  
  beda1b76
- update dailydialog: remove space before punctuation and replace " ’ " with "'" · 59fe66c5
  zqwerty authored 3 years ago
  
  59fe66c5
May 06, 2022
- Merge branch 'unified_dataset' · eee9a0a3
  zqwerty authored 3 years ago
  
  eee9a0a3
- update dailydialog preprocess.py · 0ef21cf8
  zqwerty authored 3 years ago
  
  0ef21cf8
- add dailydialog · 7bcb6096
  zqwerty authored 3 years ago
  
  7bcb6096
- Merge branch 'unified_dataset' · 052b8c5f
  zqwerty authored 3 years ago
  
  052b8c5f
- add dailydialog · 697ab75f
  zqwerty authored 3 years ago
  
  697ab75f
May 05, 2022
- Merge pull request #59 from ConvLab/pre-training · 69ba479e
  zhuqi authored 3 years ago
  69ba479e
- update base model scripts · 653b6238
  zqwerty authored 3 years ago
  
  653b6238
- Merge branch 'master' into pre-training · efeb62b9
  zqwerty authored 3 years ago
  
  efeb62b9
- Merge pull request #58 from ConvLab/nlu · f7fe7b49
  zhuqi authored 3 years ago
  f7fe7b49
- fix bug in bertnlu nlu.py: spacy tokenizer · d71de0fa
  zqwerty authored 3 years ago
  
  d71de0fa
Apr 30, 2022
- add dst fewshot shell · 3e77b6dc
  zqwerty authored 3 years ago
  
  3e77b6dc
Apr 29, 2022
- update run_seq2seq: add early stopping. update dst script to support fewshot learning · a26a42b2
  zqwerty authored 3 years ago
  
  a26a42b2
- run nlg fewshot · 74d2408e
  zqwerty authored 3 years ago
  
  74d2408e
Apr 27, 2022
- Merge branch 'master' into pre-training · f7190676
  zqwerty authored 3 years ago
  
  f7190676
- run nlu transfer exp, order=0,1,2 for few-shot(0.1,0.01) · 4f6c327c
  zqwerty authored 3 years ago
  
  4f6c327c
- add nlg fewshot, update run_seq2seq for separated truncation · 89e557bc
  zqwerty authored 3 years ago
  
  89e557bc
- extract keywords use gpt2-large and train t5-key2gen on metalwoz+sgd+tm1+tm2+tm3 · aee0a26e
  zqwerty authored 3 years ago
  
  aee0a26e
Apr 26, 2022
- Update vector_binary.py · 8b1c3fd5
  Christian Geishauser authored 3 years ago
  8b1c3fd5
Apr 22, 2022
- add t5nlu few-shot script, 1% multiwoz21: acc 54, f1 72.1; 10% multiwoz21: acc73.3, f1 84.6 · df0c81b0
  zqwerty authored 3 years ago
  
  df0c81b0
Apr 21, 2022
- Merge pull request #57 from ConvLab/nlg-scgpt · f69b60bc
  zz-jacob authored 3 years ago
  f69b60bc

Admin message